Graduation Of Wim Kleermaker: Measuring And Predicting Wear In Impellers

Wim Kleermaker during the practical phase of his graduation in dredging
Wim Kleermaker during the practical phase of his graduation in dredging

Yesterday, Wim Kleermaker graduated at the TU Delft on a research project he conducted on our slurry test circuit at Damen Dredging Equipment. Specifically, he was investigating the wear behaviour in our dredge pumps. The noteworthy aspect of this project, was that Wim was supervised by our colleague Suman Sapkota. As long time readers in the audience might remember Suman was my own pupil some years ago1.

Example of a worn impeller at the Damen Dredging Experience
Example of a worn impeller at the Damen Dredging Experience

Wear is a very common process in the dredging industry and one of the main cost factors in a project2. It is beneficial to know the amount of wear to expect in a certain condition and be able to predict the budget to reserve for this nuisance. This is only possible when we as a manufacturer will be able to predict the wear rate and pattern can provide the information to the operator for his estimates. We do have historical data that will allow us to provide a ball park figure, but a more analytical approach might assist us in particular unusual cases. Furthermore, it will also provide us insight in the impact of certain design decisions for the wear performance of a certain pump design. For Wim’s graduation, he had to approach this academically: come up with a simulation model and verify this with measurements.

Damen dredge pump slurry test circuit on the outfitting quay in Nijkerk
Damen dredge pump slurry test circuit on the outfitting quay in Nijkerk

The measurements were done in our slurry pump test circuit. This circuit has been highlighted a couple of posts back3. For Wim’s experiments, he used an impeller under a certain operating condition and mixture properties. Before and after a representative period, the condition of the impeller was measured and the difference is a measure of the wear experienced.

CFD result for the wear experiments
CFD result for the wear experiments

Wear (or scientifically: erosion) is related to the impact of the particles on the material surface. In order to know the kinetic energy of the particles, the flow field has to be known. As the flow fleild cannot be measured directly at the test circuit, we have to resort to Computation Fluid Dynamics. We already know of Suman’s graduation, to look for patterns in the flow lines, but Wim has extended the procedure to also quantitively estimate the related erosion.

Comparing CFD results (l) with measured erosion (r)
Comparing CFD results (l) with measured erosion (r)

Although there is only a limited amount of data available, comparing the results of the CFD estimation and the measured erosion are looking promising. This is certainly a workflow that will provide us the unique tools for engineering better pumps and assisting customers in their specific projects.
Although Wim will not join our ranks in the dredging community and pursue a different career in another interesting industry, we are sure he will be constructive and dedicated colleague at Marin.

Another master and student fighting evil forces (Credit: Star Wars)
Another master and student fighting evil forces (Credit: Star Wars)

References

  1. Graduation Suman Sapkota: Where wear parts were worn down, Discover Dredging
  2. Wear parts, Discover Dredging
  3. Student Interviews On Their Projects With Our Dredge Pump Slurry Test Circuit In Damen Nieuws, Discover Dredging

See also

Super Materials To Improve Lifetime When Your Pump Is On Acid

Severely corroded impeller next to the original wear part
Severely corroded impeller next to the original wear part

Recently I had a discussion on LinkedIn about the pump killer #2: ‘wrong material’. There I chipped in with this disaster picture1. It was an application where we provided a suboptimal material for the acid environment. The consequences were disastrous, as seen above. Luckily, we were able to identify the problem and propose a different material. Now, I want to share our experience here, also.

What was the case? A client requested a DOP for handling tailings in their facility. Tailings is fine stuff. Leftover from mining or waste water processing. We are always careful on the grain size, as these fines may interfere with the operation of the mechanical seal. With appropriate measures, they can handle them. As the grains tend to be fresh, they can be razor sharp. The erosion on the wear parts is higher than normal fine silt. Oh, and most tailings come with acid in their water.

So, for this request we proposed a material that was usually good in wear resistance and had a moderate resistance against corrosion. Casting materials can be classified for their corrosion resistance with the Pitting Resistance Equivalent Number2. This PREN can be calculated with:

PREN = Cr + 3.3Mo + 16N

Two observations to this formula. One, this is only valid for normal Chromium content materials intended as Stainless Steel. Two, it does not mention the aggression of the corrosion. The acidity is usually provided in the request for quotation. But, a catalyst for the oxidising reaction is the conductivity of the fluid. Chloric acid and sulphuric acid may have the same pH, but due to their different ion and electron content, their conductivity differ. We did not check this in the above example, with the consequences in the picture.

Is increasing the chromium content in the wear alloy a solution to this corrosion problem? Mwah, moderately. Alloys like stainless steel profit from the above approach. But, wear materials use their Chromium for generating carbides. Those are the particles we require for the wear resistance. The Chromium provided is than not available for corrosion resistance. e.g. White cast iron with 3%C and 21%Cr will only have about 6% of Chromium to be used in the PREN. For white cast irons, it is better to use the following graph3 to find their corrosion resistance.

Corrosion Properties of Cast Iron Ball Materials in Wet Grinding. (Credit: Corrosion Feb 1992)
Corrosion Properties of Cast Iron Ball Materials in Wet Grinding. (Credit: Corrosion Feb 1992)

If corrosion is such an issue, why don’t we use Stainless Steel? Well, there you bite yourself in the tail. Stainless Steel in itself is relatively soft. It would have the same wear index of normal construction steel. By definition a Wear Rate Index of 1. For the sharp tailing material, that would be disastrous in itself. But, let’s play along. The stainless steel derives it’s corrosion resistance from the Chromium as explained before. Chromium’s trick is to generate a clear closed patina layer of Chromium Oxide protecting the underlaying material. In dredging conditions, the particles damage the protecting patina forever exposing fresh base material for more erosion and corrosion. In the end, the wear is accelerated and part life decreases dramatically.

Accelerated erosion process under corrosive conditions
Accelerated erosion process under corrosive conditions

Back to the pictured example, we expected some corrosion, but did not expect the higher conductivity. So, after three weeks, the client noticed a sudden los of performance. The leading edge of blades and the hub shroud were completely eaten away. As long as the trailing edge was there, it generated head. A single stone hit severed the front of the impeller from the hub and we received the above disaster picture. After damage evaluation, we sent a CW250 impeller and that one lasted.

A corrosion resistant DOP working in an acid tailings pit
A corrosion resistant DOP working in an acid tailings pit

References

  1. Pump killers: How to fight the 13 most common centrifugal pump failures? Number 2., Jos Overschie
  2. Pitting resistance equivalent number, Wikipedia
  3. Corrosion Properties of Cast Iron Ball Materials in Wet Grinding, Corrosion

See also

Presenting Pump Power Peculiarities, Playing With Pumps And Pipes

Pump power exhibit at the Damen Dredging Experience
Pump power exhibit at the Damen Dredging Experience

Hej kära läsare, jag vill ta dig till ett land långt borta, för länge sedan. Min älskade Sverige.

In 1996, I started my graduation with Skanska1 in Sweden. They had a project to clean up a lake2 with an auger dredge. The auger was not performing and they asked the Delft University of Technology to investigate the problem and write a report with my solution. Off, I went to Växjö and spent a year on a dredge. During my reconnaissance of the project in the first week, I noticed that the flow in the pipe line was very slow and the motor was hardly working at full speed. As an innocent student, I asked where they were pumping the material to. ‘Oh, through 7 kilometre of pipe and 30 meter up into the hills.’ They were lucky it was such a fluid material and did not settle at such a low velocity. I then proposed they should buy a booster station to increase production3, as I could not see anything wrong with the auger. ‘No, no. It has to be the auger and the engine is strong enough; you see, there is no power required.’

Clean up project at Södra Bergundasjön near Växjö
Clean up project at Södra Bergundasjön near Växjö

That was the first time I saw the slow flow fallacy at work. Intuitively you would expect, that a long pipeline would require more power to transport the mixture than a short pipe line. This is exactly what this exhibit is trying to visualise. Water can be pumped through either the short or long pipe. From the lines on the tank wall, you can read that the output velocity of the fluid is about 1.5 m/s. In the vertical pipe, the delivery pressure is indicated. Multiplying pipe velocity and fluid pressure results in the actual power in the pipe line. The pump has to provide this power, by converting electrical power to mechanical power and eventually fluid power. On the display at the left of the buttons, the consumed electrical power can be read.

Discharge capacity through the short pipe line
Discharge capacity through the short pipe line

When you select the short pipe line, you have to notice the higher flow velocity and the required power at the display. Switching over to the longer pipeline, you will notice a drop in velocity. Due to the longer pipe, the fluid experiences more resistance. For the same pressure, the flow will be lower. Consequently, the power consumption will be lower also! This is exactly according to the theory. A pump at a lower capacity will consume less power, even if the pressure rises slightly.

Discharge capacity through the long pipe line
Discharge capacity through the long pipe line

Off course, the delivered mixture will be less, on the longer line. You might increase the speed of the pump to have more pressure. And indeed, that would require more power. But there is a maximum speed on the pump drive. Same for a very short pipe. You might end up below the idle speed of the diesel engine. Be careful in your project layout that you do take into account this viable operating range for the length of the pipe line. A longer pipeline might require a booster station for increased production. Conversely, a shorter pipe line might be chosen with a smaller diameter for increased resistance and lower power consumption, while keeping the operating point of the dredge pump near the Best Efficiency Point.

Graphical explanation of the power consumption for identical pump speeds
Graphical explanation of the power consumption for identical pump speeds

And the Swedish dredge on 7 km of pipeline? It turned out, it was not a technical problem. They had no hurry. The contractor was hired per week on an open contract…

Auger dredge 'Detritus'
Auger dredge ‘Detritus’

References

  1. Welcome to Skanska, Skanska
  2. Södra Bergundasjön, Wikipedia (Swedish)
  3. Damen booster station, Damen

See also